matrix subset problem with factors

On Feb 20, 2019, at 4:23 AM, ??? ( ??? / rIsHi ) <rishi.dasroy at gmail.com> wrote:

Hi All,

I like to report this bug related to matrix subset by rownames when passed
as factors. Now factors are may not be safe to use but then it should
generate a warning message. Since many time we use values returned by some
packages as factor to subset a matrix and which may result in a wrong
calculation.

I wish if "factor" is not expected in matrix operation then it should throw
an error/warning message.

Below are the codes to reproduce it.

x <- matrix(1:9, nrow = 3, dimnames = list(c("X","Y","Z"),

c("A","B","C")))

rNames <- as.factor(c("X","Z"))
# As some functions from different packages return factors and which

could be overlooked

rNames

[1] X Z
Levels: X Z

x[rNames,]

 A B C
X 1 4 7
Y 2 5 8

## The intended matrix should return X and Z rows instead of X and Y

sessionInfo()

R version 3.4.1 (2017-06-30)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 14.04.5 LTS

Matrix products: default
BLAS: /usr/lib/atlas-base/atlas/libblas.so.3.0
LAPACK: /usr/lib/lapack/liblapack.so.3.0

locale:
[1] LC_CTYPE=en_GB.UTF-8       LC_NUMERIC=C
[3] LC_TIME=en_GB.UTF-8        LC_COLLATE=en_GB.UTF-8
[5] LC_MONETARY=en_GB.UTF-8    LC_MESSAGES=en_GB.UTF-8
[7] LC_PAPER=en_GB.UTF-8       LC_NAME=C
[9] LC_ADDRESS=C               LC_TELEPHONE=C
[11] LC_MEASUREMENT=en_GB.UTF-8 LC_IDENTIFICATION=C

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base

loaded via a namespace (and not attached):
[1] compiler_3.4.1

With regards
Rishi Das Roy

Hi,

I get the same behavior in R 3.5.2 on macOS.

Others may feel differently, but I am not so sure that this is a bug,
as opposed to perhaps the need to clarify in ?Extract, that the
following, which is found under Atomic vectors:

"The index object i can be numeric, logical, character or empty.
Indexing by factors is allowed and is equivalent to indexing by the
numeric codes (see factor) and not by the character values which are
printed (for which use [as.character(i)])."

also applies to the indexing of matrices and arrays.

Since matrices and arrays in R are vectors with 'dim' attributes, the
behavior is essentially consistent as described above.

Thus, perhaps just add the second sentence above or similar wording to
the section for Matrices and arrays.

Regards,

Marc Schwartz

On Feb 20, 2019, at 4:23 AM, ??? ( ??? / rIsHi )

<rishi.dasroy at gmail.com> wrote:

Hi All,

I like to report this bug related to matrix subset by rownames when

passed

as factors. Now factors are may not be safe to use but then it should
generate a warning message. Since many time we use values returned by

some

packages as factor to subset a matrix and which may result in a wrong
calculation.

I wish if "factor" is not expected in matrix operation then it should

throw

an error/warning message.

Below are the codes to reproduce it.

x <- matrix(1:9, nrow = 3, dimnames = list(c("X","Y","Z"),

c("A","B","C")))

rNames <- as.factor(c("X","Z"))
# As some functions from different packages return factors and which

could be overlooked

rNames

[1] X Z
Levels: X Z

x[rNames,]

 A B C
X 1 4 7
Y 2 5 8

## The intended matrix should return X and Z rows instead of X and Y

sessionInfo()

R version 3.4.1 (2017-06-30)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 14.04.5 LTS

Matrix products: default
BLAS: /usr/lib/atlas-base/atlas/libblas.so.3.0
LAPACK: /usr/lib/lapack/liblapack.so.3.0

locale:
[1] LC_CTYPE=en_GB.UTF-8       LC_NUMERIC=C
[3] LC_TIME=en_GB.UTF-8        LC_COLLATE=en_GB.UTF-8
[5] LC_MONETARY=en_GB.UTF-8    LC_MESSAGES=en_GB.UTF-8
[7] LC_PAPER=en_GB.UTF-8       LC_NAME=C
[9] LC_ADDRESS=C               LC_TELEPHONE=C
[11] LC_MEASUREMENT=en_GB.UTF-8 LC_IDENTIFICATION=C

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base

loaded via a namespace (and not attached):
[1] compiler_3.4.1

With regards
Rishi Das Roy